Previously we had a recipe for the specific solution type that we implemented. But for some reason, this broke du to sanges of SciML. So to not have such issues again, let's just use their plot recipe! To do this we just needed to implement some recipes for the Gaussian type that we use in the solutions, and make sure that the solution interpolation etc is doing what it should do.
This turned out bigger than I thought: To properly use the SciMLBase plot recipe, I should also use the SciMLBase solution interpolation. So doing this is what's left to do for this PR.
EDIT [2024-01-24]:
[ ] Add whatever we do here to the documentation page on how we interact with the OrdinaryDiffEq internals!
[x] Maybe make the solution object actually have Gaussians as sol.u? And then the mean solution object has the means?
Previously we had a recipe for the specific solution type that we implemented. But for some reason, this broke du to sanges of SciML. So to not have such issues again, let's just use their plot recipe! To do this we just needed to implement some recipes for the
Gaussiantype that we use in the solutions, and make sure that the solution interpolation etc is doing what it should do.This turned out bigger than I thought: To properly use the SciMLBase plot recipe, I should also use the SciMLBase solution interpolation. So doing this is what's left to do for this PR.
EDIT [2024-01-24]: